Setting up a GCC in India: the definitive 2026 playbook

India is no longer a back office story. In 2026 it is the single largest concentration of Global Capability Centres in the world, hosting over 1,800 centres employing more than 2 million people and contributing close to 1.6% of national GDP. Every Fortune 500 finance team, every top 20 global bank, every major pharma, every hyperscaler and every premium engineering OEM now runs material capability out of India. The question for a global CFO or COO in 2026 is no longer whether to set up an India GCC. It is how to set one up that survives the next three macro cycles, hires the right people, sits in the right city, structures the right entity, and produces compounding strategic value rather than commodity cost arbitrage. This guide answers that question end to end.

1. Why India, and why now

Three forces have converged in 2026 to make India the default GCC destination. First, the global cost of senior technical and finance talent in the US, UK, Eurozone and Australia has risen 35 to 55% since 2020, while the India differential has held at 65 to 75% lower fully loaded cost for equivalent roles. Second, the depth and seniority of Indian talent has crossed a threshold: there are now over 180,000 product managers, 90,000 staff and principal engineers, 60,000 chartered accountants with US GAAP experience, and 40,000 actuaries and quants based in India. Third, the political and regulatory environment for inward GCC investment has stabilised: the Companies Act, the GST regime, the Special Economic Zone rules under the IFSCA framework, and the transfer pricing safe harbour rules have all settled into predictable, defensible postures.

The result is a market in which a 200 FTE GCC built in 2026 will, on a fully loaded ten year DCF basis, deliver 3.2 to 4.8x the value of the same headcount retained onshore in most Western markets. That is not arbitrage. That is structural.

2. The four entity structures and how to choose

Every India GCC sits in one of four legal structures. Choosing wrong adds 6 to 12 months of regret and 15 to 25% of avoidable cost.

• Wholly Owned Subsidiary (Private Limited): the default for 80% of GCCs. Full control, full IP ownership, cleanest transfer pricing posture, easiest exit. Incorporation in 3 to 5 weeks.
• Limited Liability Partnership (LLP): occasionally used by professional services firms and small captives. Lower compliance, but capital infusion and FDI rules are clumsier. Avoid for anything above 100 FTEs.
• Branch Office: rare in 2026. Restricted activity list, RBI approval needed, no separate legal personality. Use only when the parent specifically wants no India entity on its balance sheet.
• GIFT City IFSC Unit: the right answer for offshore fund administration, aircraft leasing, ship leasing, and tax-neutral booking of cross-border financial services. Not a replacement for a mainland GCC; a complement.

For 9 out of 10 enterprises, the right answer is a Wholly Owned Subsidiary structured as a Private Limited Company under the Companies Act 2013, with the parent holding 99.99% and a single resident director or nominee holding 0.01% to satisfy the resident director requirement under Section 149(3).

3. The transfer pricing decision that shapes everything

The single most consequential decision in setting up an India GCC is the transfer pricing model. It determines your tax exposure, your audit risk, your ability to attract senior leadership, and your strategic optionality for the next ten years. There are three real choices in 2026.

• Cost plus markup (typically 12 to 18% on operating cost): the dominant model. Predictable, defensible under the safe harbour rules for IT and ITeS services, audit friendly. Best for delivery centres, shared services, and most engineering work.
• Profit split: rare, used where the India centre creates genuinely co-developed IP with the parent. Higher tax, but unlocks the ability to attract entrepreneurial leadership with equity-linked outcomes.
• Berry ratio: niche, used for distribution and low-value-added services. Avoid for true capability centres.

Most 2026 GCCs land at a cost plus 15% model and apply for safe harbour where eligible (turnover below INR 200 crore for ITeS, INR 200 crore for software development). Above those thresholds, an Advance Pricing Agreement with the Indian tax authorities is the prudent path. APAs now take 18 to 24 months to conclude; start the conversation in parallel with incorporation, not after.

4. City selection: the four-quadrant model

India has four tier-1 GCC cities, and the choice between them is rarely a tie. Each city has a function it does best, a function it does adequately, and functions it should be avoided for.

• Bangalore: best for AI and ML platform engineering, deep product engineering for SaaS, semiconductor design, cybersecurity engineering, and venture-grade tech leadership. Avoid for high volume accounting or BFSI middle office.
• Pune: best for Accounting, Controllership, FP&A, mechanical and electronics engineering, Japanese and German captives, and any function that needs to feel boringly stable for seven years. Avoid for capital markets or AI research at the frontier.
• Mumbai: best for capital markets operations, investment banking middle office, insurance and actuarial, treasury and FX, regulatory reporting, and any work that benefits from physical proximity to the RBI, SEBI, NSE, BSE and the global bank India HQs in BKC. Avoid for cost-sensitive volume work.
• Hyderabad: best for pharma and life sciences GCCs, biopharma data science, US BFSI back office, healthcare payer operations, and large hyperscaler engineering centres. Avoid for European or Japanese captives where Pune wins on cultural fit.

For multi-function GCCs above 500 FTEs, the dual-city pattern (Pune plus Mumbai, or Bangalore plus Hyderabad) increasingly outperforms a single-city build. It diversifies talent, cost, climate, and political risk, and lets each function sit in the city built for it.

5. The 180 day execution plan

A well-run India GCC goes from board approval to first 50 hires in 180 days. The phases are sequential and unforgiving; compressing any of them tends to add 90 days at the end.

• Days 1 to 30: legal entity incorporation, PAN, TAN, GST registration, professional tax, Shops and Establishment registration in the chosen state. Open INR and USD bank accounts. File FC-GPR with RBI within 30 days of share allotment.
• Days 31 to 60: appoint the India MD or Head of GCC. This is the single most important hire. It must be a 15+ year operator who has built a GCC before, not a functional specialist promoted into the role.
• Days 61 to 90: real estate selection. Shortlist 4 to 6 Grade A buildings in the chosen submarkets. Sign a Letter of Intent. Negotiate a 5+5 year lease with a 24 to 36 month lock-in, escalation capped at 5% every three years, and a fit-out period of 90 days rent-free.
• Days 91 to 120: hire the leadership team (Heads of HR, Finance, Technology, the 3 to 5 functional leads). Set up payroll, group medical insurance, provident fund, gratuity and ESIC. Sign master service agreements with the parent for the transfer pricing arrangement.
• Days 121 to 150: fit out the office. Hire the first 25 individual contributors. Run the first parent-India offsite to lock in operating rhythm.
• Days 151 to 180: hire the next 25 individual contributors. Go live with the first deliverable. Run the first quarterly business review with the parent.

6. The hiring playbook that actually works in 2026

Hiring in India in 2026 is harder than at any point in the last decade. The top quartile of talent has 4 to 7 active offers at any time, attrition in the first 12 months runs 22 to 28% in Bangalore and 16 to 20% in Pune, and counter offers from the existing employer are now standard at 25 to 35% increments. The captives that win are the ones who treat hiring as a craft, not a transaction.

• Source from the second tier of every target company, not the headline names. The third senior manager at a tier-1 bank is often more effective and more loyal than the second VP at a tier-2 bank.
• Pay at the 65th to 75th percentile of the market, not the 90th. Above the 75th percentile you attract candidates who are optimising for compensation, not for the opportunity.
• Build the employer brand on outcomes, not perks. Indian senior talent in 2026 cares about scope, ownership, exposure to global leadership, and learning from the parent. Free meals and pet-friendly offices stopped being differentiators in 2018.
• Make every offer in 7 days from first interview. Drag the timeline and you lose the candidate to a parent that moved faster.
• Invest in the first 90 days. A structured onboarding with weekly leadership exposure cuts first-year attrition by 35 to 45%.

7. The compensation and benefits architecture

Indian compensation has converged on a structure that almost every GCC follows in 2026. Base salary is 55 to 65% of fully loaded cost. Variable pay is 12 to 20% of base, linked to a mix of parent global outcomes and India centre outcomes. Long-term incentive is offered to the top 5 to 15% of the centre, typically as parent listed company RSUs vesting over 4 years. Group medical insurance covers the employee, spouse, two children and both sets of parents up to INR 10 lakhs per family. Provident fund is 12% of basic from employer and 12% from employee. Gratuity is statutory at 15 days per year of service after 5 years. The total fully loaded cost runs at base x 1.42 to 1.55 depending on city and seniority.

8. The technology stack a 2026 GCC inherits

The parent stack flows down by default, but the India GCC almost always layers four India-specific decisions on top. First, the identity and access management: enterprise IDP federated to the parent, with India-specific role definitions and quarterly access reviews mandated by the Digital Personal Data Protection Act 2023. Second, the collaboration stack: Microsoft Teams or Slack standardised across the centre, with a parent-India bridge for asynchronous handoff. Third, the data residency posture: any personal data of Indian data subjects stays in India under the DPDPA, which means a dedicated India tenant for HRMS and customer support tooling. Fourth, the developer infrastructure: GitHub or GitLab Enterprise, a parent-managed cloud account with India-region resources, and a clear policy on what data can and cannot leave India.

9. The five most expensive mistakes

Almost every painful India GCC story traces back to one of these five mistakes, made in the first six months and paid for over the next five years.

• Hiring the India MD too late or too cheaply. This person sets the centre culture, the hiring bar, and the parent relationship. Pay for the best one you can find. Hire before you sign the lease, not after.
• Choosing the city to match the cheapest rent rather than the function. Pune does not do AI research as well as Bangalore. Bangalore does not do BFSI middle office as well as Mumbai. The rent saving disappears in the first year of attrition.
• Underinvesting in the parent-India operating rhythm. Captives that fail almost always fail because the parent treats them as a vendor, not as a colleague. A weekly leadership cadence, a quarterly in-person offsite, and shared OKRs are non-negotiable.
• Building too much real estate too fast. Take the smaller floor with an option to expand. Empty desks compound into culture problems and procurement scrutiny.
• Treating compliance as a back office function. The Companies Act, the GST, the DPDPA, the labour codes, the SEZ rules: each one can stop the centre cold if mishandled. Hire a senior India CFO or finance controller in month two, not month eight.

10. The 24 month value curve to plan for

A well-built India GCC follows a predictable value curve. In months 1 to 6 the centre burns cash and produces nothing visible to the parent. In months 7 to 12 the first 50 FTEs go live and the centre starts producing 60 to 70% of the equivalent onshore output at 30 to 35% of the cost. In months 13 to 18 the centre crosses 150 FTEs and starts producing work the parent could not have produced at all because of talent scarcity onshore. In months 19 to 24 the centre starts contributing strategic IP, owns end to end products or processes, and the conversation in the parent moves from 'how much does India save' to 'what should India own next.' Plan the parent communication, the headcount ramp, and the capability roadmap around this curve. The CFOs who win are the ones who tell this story to their boards in month 1, not in month 24.

11. The closing thought for a global CFO or COO

An India GCC is not a cost decision. It is a capability decision wearing a cost saving disguise. Built well, it becomes the second engine of the enterprise: the place where new capabilities are tested, where global talent shortages are absorbed, and where the parent's strategic optionality is multiplied. Built badly, it becomes a permanent line item of underperforming overhead. The difference between the two outcomes is rarely the country, the city or the rent. It is the seriousness of the first six months. Treat those six months as the most important six months of the next decade of the enterprise, and the next decade tends to look after itself.